Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - NSR

Pages: [1] 2 3
1
NSR Invest / Re: not all fields display "N/A" for missing data
« on: October 07, 2016, 03:41:54 PM »
Hi Fred,

The N/A values are there, click zoom (+) once.

It's the tolerances we have that automatically set the zoom range are confused on N/A values.

We will try better handle that. Because Lending Club has turned data on, and off, and on  there are some gaps which are tricky to handle.

Michael

2
"notes available" ... I'm thinking that means the number of notes sitting on the web site.  This counts some loans many times (day after day).

I prefer counting only new loans, so each one is counted once.

This is the number of loans we saw, new, on the APIs broken out by fractional and whole.

A loan not fully funded from day-1 would not be included in day 0

Hope that helps.


3
Anil,

1.)

Please see: http://blog.nsrplatform.com/2014/01/nickel-steamroller-2-0/ in response to not giving credit where credit is due.  This is just one example of the many.  I am also not familiar with the argument I had with Ken, if you have any references, please feel free to cite them.

2.)

Lending Club had a negative return in 2007. The total loss exceeded the the interest paid. You are reporting +1.18%
Even Lending Club confirms it was a down year: https://www.lendingclub.com/info/demand-and-credit-profile.action

Please be mindful before attacking others and calling their methodology a "POS"

Michael

If you knew the history, you wouldn't quote NSR ROI. It is POS that they copied from Lendstat. There was actually arguments between Lendstat and NSR as NSR had lifted the equation without giving credit to Lendstat. During the early age of p2p lending, the lendstat guy was trying to come up with an equation that approximated p2p lending portfolio IRR. IMO, he came up with that equation by trial and error noticing that it gave the closest number to IRR of his and few other people's portfolios. There is no theoretical basis to that equation. In retrospect, it looks like net return divided by portfolio duration where portfolio duration is approximated by interest received divided by interest rate.

I can't find it right now but Lending Club had hired a consulting company to identify proper measure for loan performance who proposed NAR in a published LC sponsored white paper.

For comparing actual portfolio performance, I suggest using any money-weighted or time-weighted return measures such as IRR. For backtesting, I suggest using NAR. If you want, you can take computing-intensive approach: create sample portfolios, account for payment timing and amount, reinvest that amount in loans available during payment events, and so on ... but return on effort is not worth it.

Did you realize that "Nickel Steamroller" (NSR) uses 0% as well?  They've been around a while and probably have analyzed this pretty well just like LC has.

NSR back testing allows you to easily see loss rates by year.  One could, for example, look at the avg of 2011-2015 loss rates, for 36A's, 36B's, ... 60F's, 60G's to try to adjust a personal portfolio.

4
Investors - LC / Re: Differing stats
« on: August 05, 2015, 03:28:05 PM »
Can you clarify "there's up to a 2% IRR difference between loan amounts on low to end scale.  NSR says no difference at all".  Maybe some screenshots to help illustrate your point?

5
P2P-Picks / Re: P2P Picks and NSR
« on: May 22, 2015, 04:43:20 PM »
That's correct.  Models are different than filter based investing.  NSR allows you to pair P2P picks with other filters though, say you only want a certain grade or another credit attributes, you can do that.  Models don't use filters in a traditional sense, but we give you the flexibility to do so with models.


6
NSR Invest / Re: NS reported AVG AGE
« on: August 13, 2014, 04:41:15 PM »
I replied to your PM, see if that solution would work and the rationale for it.

7
NSR Invest / Re: NS reported AVG AGE
« on: August 13, 2014, 11:48:40 AM »
You are correct, in the case of the back testing tools the age is defined as today - issue.   It's the most objective age calculation. The sole purpose of the age column is to give you a sense of how "old" a group of loans are.  If we use the last payment date, the purpose of the column is lost. 

If you see a group of loans with a high ROI, but young age vs high ROI and old age, the investor can reasonably be convinced the loans in the older age group will produce better risk adjusted returns.

If you use the last payment data, you now have two variables: age and repayment status.  It would make for a bad test since you aren't able to tell which of the factors is impacting the calculation.

Hope that helps.

8
NSR Invest / Re: WhereIs Debt-to-Income ratio? Not DTIw Prosper Loan
« on: June 26, 2014, 10:57:36 AM »
DTIw Prosper Loan is the DTI including the amount of money the borrower is requesting.

Take a look at: https://api.prosper.com/Listings - this is the data source we use, it provides much richer information than the data exports.

This is all the data Prosper Makes Available via the API, DTI is not one of them (not sure why).   Since Prosper groups incomes into buckets (i.e. $0-24,999) it's not really possible to get the values from the data given.

Since the API does not offer this information, we don't/can't use it, the filters you save are then used for auto investing. We would have filters that filter on data that does not exists!

9
NSR Invest / Re: lending club data for completed loans
« on: June 10, 2014, 11:34:18 PM »
This should be working now.

10
NSR Invest / Re: lending club data for completed loans
« on: June 10, 2014, 01:36:48 PM »
Pretty simple explanation, I committed a change that referenced the wrong column which over stated interest payments which in term overstated returns.  The bug only existed for a day or so. You just happened to catch it.  It's been corrected.

If you are curious about auditing returns on NSR, I recommend you check out: http://www.nickelsteamroller.com/#!/stats/roi which explains the process by which returns are calculated in depth.

11
NSR Invest / Re: Neat Stuff!
« on: February 16, 2014, 10:53:21 PM »
That is missing for now. We will add a per note view as well though. That will be important to analyzing duplicates as well.

12
NSR Invest / Re: How to log in on the new version?
« on: February 16, 2014, 10:51:51 PM »
We just applied a patch for  Safari 4.0. It should work now.

13
NSR Invest / Re: How to log in on the new version?
« on: February 10, 2014, 02:35:44 PM »
Is this still an issue? We added support for placeholders on old browsers.   We actually didn't test on safari 4.0. If I am reading this correctly that browser is for years old and hasn't been upgraded at all since OSX tiger.

I would recommend upgrading your browser to a current version if the issue still persists.

14
NSR Invest / Re: Neat Stuff!
« on: February 10, 2014, 02:08:42 PM »
You can't end a post like that and not expect a follow up!  What other things would be nice to have. We're working on release tasks for the next release coming up in a few weeks. If you have some ideas, feel free to post them or use the contact link on the site. I'll see how we might be able to get them in.

15
NSR Invest / Re: How to log in on the new version?
« on: February 05, 2014, 10:12:10 AM »
What browser are you using?  We're using some newer technology that may not work with IE7.  We've tested on IE9+, Chrome, Firefox and Safari.


Pages: [1] 2 3